putilov_denis - stock.adobe.com


AI boosts efficiency in data management

AI can automate tasks across every aspect of the data management process, enabling data teams to focus on models, not labeling and graphing.

Senior leadership and data teams are turning to AI-guided data management to help manage their ever-growing library of data, become more efficient and reduce costs.

Data is one of the most significant assets for an organization, and it continually increases in volume and importance. The data management process -- cleaning, extracting, integrating, labeling and organizing data -- is expensive and labor-intensive and presents many challenges. Data teams can invest in tools that use AI to automate aspects of the data management process.

"Adoption is in the early stages right now, but executives in all companies are prioritizing investments in data and AI technologies," said Zakir Hussain, EY Americas data leader. "Therefore, I believe that adoption will accelerate very fast."

Automating cumbersome parts of the data management process with AI can reduce the amount of manual effort required for tasks, speed up completion, increase the accuracy of outputs and cut costs. Demand for data management will continue to grow as data quality is a necessity for successful AI use.

Data management process

Data management involves many steps, which fall into several categories:

  • Data intake and storage. Enterprises must collect, process, validate and store data. Data intake includes integrating structured, unstructured and semi-structured data from multiple internal and external sources.
  • Data quality. Management also sets quality standards for the data. Quality needs generally vary based on data use and teams must maintain data to meet quality standards.
  • Data governance. An organization's data collection and use must adhere to privacy and security regulations and standards.
  • Data transformation. Accumulated data must be ready and available for use by the organization. Quality data sets are imperative as input for machine learning applications and other automated systems, as well as for data-driven decision-making.

For years, data teams have employed various technologies to support data management tasks. Many still struggle to scale their operations as the amount of data increases and enterprise data use rapidly expands.

Automate data management using AI

AI helps streamline data reporting and simplify data interpretation. It can support metadata management. AI can also automate tasks such as data modeling, making access policies and schema rule generation.

Additionally, AI can aid with data classification, cataloging, integration, quality and security. It can also enrich master data, which is the core, nontransactional data that describes the components of a business and its activities.

"Any augmentation to make these steps faster and easier for data scientists is of value, because the less time they spend on these steps, the more time they can spend on models," said Sumit Agarwal, VP analyst with the research firm Gartner.

AI's application to data management tasks includes the following:

  • Data cleansing and quality. Cleansing fixes issues in a data set, such as incorrect, incomplete or duplicate data. AI-enabled tools can detect and correct duplicate, missing and inconsistent data. AI -- specifically generative AI (GenAI) -- can aid processes further by analyzing data quality requirements, creating data validation rules and flagging errors, Hussain said.
  • Cataloging. Data cataloging is the process of inventorying all of an organization's data. It can include collecting, storing and labeling data. "We do data cataloging for governance and literacy, but it has always been a struggle," said Matt Sweetnam, chief architect at the consultancy AHEAD. AI-guided software can replace the admin's task to label, identify and quantify data, and do so as it comes in, he said.
  • Labeling. Labeling, sometimes called data annotation, is an essential part of readying data for enterprise use, particularly in machine learning models. It involves the identification and labeling of raw data regardless of type. Labeling is a large amount of work, particularly when processing images as well as when scaling up the volume of raw data to be ready for use, Argawal said. The implementation of AI to annotate data is practically non-negotiable. "It improves accuracy and the amount of time it takes to get these jobs done," he said.
  • Visualization. AI-based tools can graph the relationships within the data and weight data within a 3D display. Visualization can help teams better understand the data.
  • Augmentation. Some tools have AI-supported data augmentation. AI can automate the data enrichment process and create synthetic data to expand existing datasets, in addition to offering augmented data discovery.
Any augmentation to make these steps faster and easier for data scientists is of value, because the less time they spend on these steps, the more time they can spend on models.
Sumit AgarwalVP analyst, Gartner

Data management tools market

Demand for data management technologies shows significant growth. The global enterprise data management market totaled $89.34 billion in 2022, and could grow at a compound annual growth rate of 12.1% through 2030, according to a report by Grand View Research.

Some of that growth comes from data teams bringing in new tools that incorporate AI. Data teams typically look for data management tool vendors that incorporate AI into their products and platforms. However, some teams are building their own models to address their organization's unique needs.

Data teams can expect a growing number of AI options to support their work in the future, according to market experts. For example, AWS is incorporating AI into integrations between database services to eliminate time-consuming extract, transform and load work. Informatica uses AI-powered governance tools to automate what data a user can access. SAP is working on data quality and access improvements by incorporating GenAI into SAP DataSphere. These vendors are among the many prominent players in the enterprise data management space, according to Grand View Research's report.

"Every tool vendor is trying to figure out how to inject this technology to make their product better," Sweetnam said.

Mary K. Pratt is an award-winning freelance journalist with a focus on covering enterprise IT and cybersecurity management.

Dig Deeper on Data management strategies

Business Analytics
Content Management